Determinants of Spatio-Temporal Patterns of Cybercrimes in the USA: Implications for Cybersecurity Personnel Resource Allocation

Funder: DHS Center of Excellence in Criminal Investigation and Network Analysis
Study Period: January 2024 - June 2024

Objectives

Methodology

Spatio-temporal Regression: We leveraged a geographically and temporally weighted regression (GTWR) approach to statistically study the various determinants influencing the evolving spatio-temporal scales of cybercrime across the United States.

Cybercrime Data

Cybercrime data comes from the Privacy Rights Clearinghouse (PRC).

Socio-Economic Data

The data collected comes from the U.S. Bureau of Economic Analysis (BEA) and the American Community Survey (ACS) from the U.S. Census Bureau.

Technological Metric

This study considers a technological variable as the percentage of households with internet acces.

Backtesting

Backtesting results for each state as the percent difference between the observed and forecasted data for 2022.

Cybercrime Severity in 2024

The forecasted number of compromised records as the severity of cybercrime for each state in the year 2024.

Cybercrime Severity by FBI Field Office

Forecasted number of records compromised for each FBI main field office for year 2024.

Cybersecurity Personnel Resource Allocation

The minimum and maximum bands of the number of records compromised for each FBI agent in corresponding FBI main field offices in 2024. The blue dots represent the expected number of compromised records.

Team

  • Stefano Chiaradonna

    Ph.D. candidate in Applied Mathematics at Arizona State Univeristy, School of Mathematical and Statistical Sciences. Received his bachelor's degree from Benedictine University in mathematics. Interests include cyber risk analysis and critical infrastructure resilience.

  • Cody Delos Santos

    Ph.D. candidate in Applied Mathematics at Arizona State Univeristy, School of Mathematical and Statistical Sciences. Received master’s degree from SFSU in mathematics. Interests include machine learning, big data, and analysis of crime patterns.

  • Petar Jevtić, Ph.D. (PI)

    Associate Professor at Arizona State Univeristy, School of Mathematical and Statistical Sciences. Passionate about mathematical conceptualizations of risk. Special interests in emerging risks such as cyber risk and climate risk.

  • Kyran Cupido, Ph.D.

    Assistant Professor in the Department of Mathematics and Statistics at St. Francis Xavier University in Canada. Research focus on actuarial domains of longevity risk and P&C insurance.